Application of satellite remote sensing in monitoring dissolved oxygen variabilities: A case study for coastal waters in Korea

Abstract

Dissolved oxygen (DO) is one of the critical parameters representing water quality in coastal environments. However, it is labor- and cost-intensive to maintain monitoring systems of DO since in situ measurements of DO are needed in high spatial and temporal resolution to establish proper management plans of coastal regions. In this study, we applied statistical analyses between long-term monitoring datasets and satellite remote sensing datasets in the eastern coastal region of the Yellow Sea. Pearson correlation analysis of long-term water quality monitoring datasets shows that water temperature and DO are highly correlated. Stepwise multiple regression analysis among DO and satellite-derived environmental variables shows that the in situ DO can be estimated by the combination of the present sea surface temperature (SST), the chlorophyll-a, and the SST in the month prior. The high skill score of our proposed model to derive DO is validated by two error measures, the Absolute Relative Error, 1-ARE (89.2%), and Index of Agreement, IOA (78.6%). By applying the developed model to the Moderate Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imaging Radiometer Suite (VIIRS) products, spatial and temporal changes in satellite-derived DO can be observed in Saemangeum offshore in the Yellow Sea. The analysis results show that there is a significant decrease in estimated DO between summer of 2003 versus 2012 indicating summer coastal deoxygenation due probably to the Saemangeum reclamation. This study shows the potential capability of satellite remote sensing in monitoring in situ DO in both high temporal and spatial resolution, which will be beneficial for effective and efficient management of coastal environments. Keywords: Remote sensing, Satellite, Dissolved oxygen, Multiple regression, Yellow Sea